Parent/child dimensions in Jasper Analysis

Here I want to share my experience of using a dimension based on tree-like table. As database engine I used Oracle.

As known tree-like table is a table that has pair of columns, usually called as ID and PARENT_ID. Values of PARENT_ID column in children rows reference to values of ID column of their parents. Usually tree's element without any parent is called root, an element that has children is called node and an element without any children is called leaf.

Simulation of outer joins in Jasper Analysis

JasperReports Server (Pentaho Mondrian behind the scenes) doesn't use outer joins when parses schema file to build analysis view. Today, at least.

Here I am describing the one of possible ways to change JasperServer behaviour as though it can handles outer joins. We don't have to make changes in any configuration, the solution just encloses manipulations with SQLs have used in schema definition.

Disable cache in Jasper Analysis

As known JasperServer uses Pentaho Mondrian as base of Jasper Analysis component. Mondrian has elaborate cache which truly increases performance during using analysis views.

But  when we have not large database and need realtime OLAP the cache becomes a wall. And we want to turn off it.

I wanted to achive it using some properties in configuration and not using any Java code.